Toward a definition of pharmacoresistant neuropathic pain

Pharmacological treatment of neuropathic pain relies currently on evidence from large randomised controlled trials. Systematic reviews and recommendations are available, allowing clinicians to adopt a rational based on available scientific evidence when treating neuropathic pain (Attal et al., 2006; Dworkin et al., 2007). Despite the advances, a considerable number of patients do not get sufficient pain relief and improvement in quality of life (QoL) from available drugs. In evidence-based studies on pain it is customary to consider “responders” to treatment those patients that report a pain relief amounting to 30–50%. Meta-analysis (Finnerup et al., 2005) as well as experienced clinicians claim that this goal is obtained in only 30–40% of patients across diagnostic entities where all evidence-based pharmacological strategies have been tried. Currently we lack strategies on how to identify responders/non-responders and treatment approaches can best be characterised as trial and error. In the optimal scenario pathophysiological mechanisms of the pain should be fitted to drugs with a mechanism of action specifically targeting the former. At present we are devoid of detailed information on mechanisms of spontaneous and stimulus-evoked neuropathic pain which offsets a mechanism based treatment approach with available drugs (Baron, 2006; Woolf et al., 1998). Regardless of criteria defining responders, side effects before any or optimal pain relief may disqualify further treatment and may thus be a basis for pharmacoresistancy. The necessity for defining pharmacoresistant neuropathic pain was derived from the clinical observation that patients exist in whom sustained pharmacological rotation does not produce the desired pain relief or induces intolerable side effects, and hence unaltered low QoL or even worse, a decrease in QoL due to side effects being added to the pain suffering. Such ineffective treatment is also a considerable cost for society. Continued prescription of one drug after another may also add unmet expectancy, anxiety and frustration to the patients’ already difficult situation, thus contributing to their psychological derangement and lack of mobilisation of cooping strategies. Moreover, we have only scarce evidence for a combination therapy aiming at different molecular targets (polypharmacy) (e.g., Gilron et al., 2005) and no conclusions can be drawn (Cruccu, 2007), although theoretical implications have been put forth supporting drug combinations in peripheral neuropathic pain states (Hansson and Dickenson, 2005). For some neuropathic pain conditions alternative treatment approaches exist, such as electrical neurostimulation (Cruccu et al., 2007) and pharmacological intrathecal therapy with ziconotide (Williams et al., 2008). Furthermore, when pharmacological strategies fail and no other procedures are indicated measures such as cognitive-behavioural therapy and physiotherapy may be considered to improve coping strategies and bodily functions. Currently there are no established criteria for what should be considered pharmacoresistance with regard to neuropathic pain. If efforts are to be made toward solving the managing problem of patients who report lack of significant pain relief despite having tried multiple drug monotherapy, a definition of pharmacoresistant neuropathic pain is critical. We propose the following definition of pharmacoresistant neuropathic pain: “A neuropathic pain condition is resistant to pharmacotherapy when mono- or a rational combination treatment using drugs proven efficacious in RCTs fails in inducing useful pain relief from the patients/physicians point of view after an appropriate duration of treatment with adequate dosage, or if intolerable side effects occur”. In this context we need to define efficacious drugs, including polypharmacy, useful pain relief, adequate dosage and treatment duration. Unfortunately, data is lacking to support an evidence-based approach to the definition of these terms. In the interest of drawing attention to and promote studies and debate within this area we put forth the following. We suggest adopting the treatment paradigm recently put forth in European (Attal et al., 2006) and international recommendations (Dworkin et al., 2007) using drug classes that have proven efficacious in RCTs including tricyclic antidepressants (e.g., amitriptyline and nortriptyline), serotonin noradrenalin reuptake inhibitors (duloxetine and venlafaxine), alfa-2-delta ligands (gabapentin and pregabalin) and opioids (oxycodone, morphine, methadone and tramadol). Also, topical lidocaine for small areas of pain/mechanical allodynia and sodium channel blockers (carbamazepine and oxcarbazepine) in trigeminal neuralgia should be considered. The treatment paradigm allows for any peripheral neuropathic pain condition to be treated in a monotonous way although the scientific evidence mostly refers to studies on post herpetic neuralgia and painful diabetic polyneuropathy. A rational for monotonous pharmacological treatment of peripheral neuropathic pain conditions based on shared commonalities, i.e., similar pathophysiological mechanisms and overlapping clinical phenomenologies, despite multiple etiologies has recently been published (Hansson and Dickenson, 2005) and has gained support from a recent study (Attal et al., 2008). The efficacious drugs should be rotated per mechanism of action, i.e., there is no evidence currently to recommend trying two drugs with the same mechanism of action unless impossible to reach the adequate dosage or treatment duration with one of them because of intolerable side effects that the second drug is not expected to produce. Available controlled studies have only provided evidence in favour of combining gabapentin with opioids (Gilron et al., 2005; Hanna et al., 2008) or gabapentin and venlafaxine (Simpson, 2001). A rational approach would be to combine drugs from the drug classes that have proved efficacious in relieving neuropathic pain (Hansson and Dickenson, 2005). In clinical practice the usefulness of a treatment is decided on by the patient and the prescribing physician. A decision to continue medication is based on summing up the balance between the drug benefits on pain and comorbidities on one side and the adverse events on the other, i.e., the patient concludes that the quality of life is sufficiently improved. Such a decision does not necessarily correspond to the 50% or 30% pain relief often referred to in clinical trials and is probably better matched by the commonly entertained Patient Global Impression of Change (PGIC) (Farrar et al., 2001). In this context we would like to point out that a drug might act differentially on the flora of neuropathic pain symptoms and signs in a single patient (Baron, 2006). How the patient decides on the overall efficacy of a drug in such a situation is unknown but should be detailed in the course of drug evaluation. Alleviation of some symptoms/signs is not necessarily reflected in the global rating of the pain. The role of the physician in this process is to support the patient in the evaluation process, to carefully weighing pros and cons of continued treatment from a medical perspective and, in addition, to consider health economic aspects. While awaiting further studies, evidence-based data suggests that at least a moderate improvement using the PGIC-scale may be an indicator of useful pain relief (Farrar et al., 2001). The adequate dosage range is defined by RCT results of efficacious drugs. If the patient can take the highest recommended dose (i.e., side effects considered tolerable), this should be tried before considering treatment failure. Here the pharmacokinetics of the drug in combination with recommended titration time serves as a basis. To what extent the pharmacodynamics of a drug at maximally tolerated dose requires a certain time to optimally come into effect is not known. A maintenance period of 3weeks before evaluation is proposed, and longer if suggested by available RCTs. Furthermore, at regular intervals (e.g., 6months) it should be assessed by drug tapering whether a therapy is still efficacious. The required scientific evidence to put forth a non-disputable definition of pharmacoresistant neuropathic pain is not available. To improve this, a major workload is foreseen with, e.g., head-to-head studies of available drugs in different clinical diagnostic entities. In the interest of the patients we believe that the issue of pharmacoresistance in neuropathic pain is important and hence consider it useful to propose this preliminary definition and arguments to attract further discussion and development. Also, we desperately need more efficacious drugs and accumulating information about the usefulness of polypharmacy and effectiveness of non-pharmacological measures.